RFR(S): 8012322 : Tiered: CompilationPolicy::can_be_compiled(CompLevel_all) mistakenly return false

Igor Ignatyev igor.ignatyev at oracle.com
Thu Apr 18 15:11:59 PDT 2013


 > For the code to be "clear and self-documented" :) use some meaningful
 > names instead of 'i', 'j' and 'n'.
renamed

http://cr.openjdk.java.net/~iignatyev/8012322/webrev.01/

Best regards,
Igor Ignatyev

On 04/18/2013 10:02 PM, Vladimir Kozlov wrote:
> Changes are good. My only complain is variables naming:
>
> !             final int n = TIERED_STOP_AT_LEVEL + 1;
> +             for (int i = 1; i < n; ++i) {
>
> +         for (int j = 1, n = TIERED_STOP_AT_LEVEL + 1; j < n; ++j) {
>
> For the code to be "clear and self-documented" :) use some meaningful
> names instead of 'i', 'j' and 'n'.
>
> Thanks,
> Vladimir
>
> On 4/18/13 10:30 AM, Igor Ignatyev wrote:
>> Hi all,
>>
>> Please review the patch.
>>
>> Problem:
>> if method is 'not_compilable' in C1 or C2,
>> 'CompilationPolicy::can_be_compiled' will return false for
>> 'CompLevel_all' even it is compilable in another level
>>
>> Fix:
>> 1. added explicit comp_level to all calls of
>> CompilationPolicy::can_be_compiled
>> 2. in CompilationPolicy::can_be_compiled()
>>    - added assert for checking legality of comp_level
>>    - changed logic for comp_level == CompLevel_all
>>
>> Testing:
>> JPRT 'compiler/whitebox'
>>
>> webrev: http://cr.openjdk.java.net/~iignatyev/8012322/webrev.00/
>> jbs: https://jbs.oracle.com/bugs/browse/JDK-8012322


More information about the hotspot-compiler-dev mailing list